About
The development of an integrative model which examines the biological plausibility of how bilinguals recognize words from the two languages, apparently without much effort. Despite a growing interest in this issue over the last 30 years, computational models designed to deal with the intricacies of the architecture of the bilingual memory are scarce. One leading computational model that has received strong empirical support (cf. Comesaña et al., 2012, 2015) is the Bilingual Interactive Activation Model (BIA+, Dijkstra & van Heuven, 2002; Dijkstra, Miwa, Brummelhuis, Sappelli, & Baayen, 2010), which proposes an integrated lexicon with non-selective access to the two languages. Although the model incorporates orthographic, phonological, linguistic and non-linguistic context effects and a task-decision component that can account for a wide variety of empirical data, it also presents several shortcomings. For instance, it is currently not equipped to explain: a) the precise representation of identical and non-identical cognate words (equivalent translations that share form and meaning) in the bilingual memory; b) the feature/letter to word connections that could explain why cognate processing varies as a function of the position of the deviant letter (faster processing for cognates that have a deviant-letter at the end [texte-texto, in French and Spanish, respectively] than for cognates whose deviant letter is at the middle [usuel-usual]; Font, 2001; and c) the specification of the mechanisms that fit age-related effects on reading (see Li, 2002). Overcoming these limitations has been the main aim of the present work. Specifically, to develop an extension of the BIA+ model incorporating a full characterization of the connections which can address the aforementioned issues.
